Chapter 10 — The City Has No Smell

They set up a testing room right in the courtyard of the culture house, without using white walls and stable air conditioning. A row of tables is divided into four stations: odor description, temperature measurement, respiratory rate recording, and location marking. The goal is not to find the poison in an afternoon. The goal is to see if the city can observe the same phenomenon without turning one person's feelings into accusations or one machine's measurements into denials.

Huyen wrote his goal on the big board:

"Measure what the machine records, what the person perceives, and what the body reacts to at the same time. Don't infer the cause without enough data."

Nghi read the board and nodded. "The first sentence is okay. The second sentence sounds like you're scared."

"I'm scared."

"Good. People who aren't afraid of data often turn it into knives."

Tung placed three measuring boxes on the table. He has replaced the calibration box for station number 3, but has not returned the system to automatic mode. Quynh allows testing for two hours with a publicly announced spraying schedule. Van prepared the symptom chart, and Tu brought twelve descriptive cards from the odor inventory.

More people arrived than expected. Some people think this is a free health check-up. Some people want to protest against the Clean Class. There were workers who were curious because they were seeing their station data for the first time. Huyen does not prevent different purposes. She just asked everyone to choose the right table.

The station broadcasts diluted standard odor samples. Participants recorded intensity and nature. Station two measures air at three altitudes. Station three counts breathing before and after traversing a route. Station four had people mark where they felt their bodies change. Huyen reminded the groups not to tell each other the results before they finished recording.

An is in charge of the map. He wore a small name tag that did not say "patient" but "observer". A little girl asked him if he was a doctor. An shook his head.

“I draw places where breathing is difficult.”

“Does drawing know about illness?”

“No. But know how to ask.”

Tung heard it and said to Huyen: "He answered the same as you."

“No. I answered the same as him.”

The first spraying begins at 3:00 p.m. The culture house loudspeaker clearly reads the station location, time, purpose and warns that Clean Class is not a complete air purifier. Some people applauded, some said the announcement was too long. Huyen kept the same. A warning cannot be short enough for speakers and clear enough for every body.

At station one, participants received samples of the smell of organic mud before going onto the field. At station two, the gauge recorded an increased moisture layer. At station three, the average respiratory rate remained constant for the first five minutes. At station four, three people marked the area near the embankment where their throats were dry.

Huyen does not allow anyone to say "contradiction". She said those are perspectives that have not been connected by time.

Nghi stood at the food sample table, carrying two identical jars of fish sauce. One bottle was taken from her counter, one bottle was placed in the room near the spray head for ten minutes. Participants did not know which jar was which. Seven people noticed that the second bottle was "less open", four people said it was no different, one person said the second bottle had a slight plastic smell.

Nghi asked Huyen: "Is it enough to say that the fish sauce making system is broken?"

"Not yet."

“Is it enough to disable the spray system when inspecting food?”

“Enough to recommend private testing.”

Nghi nodded. “I like the word proposed more than the word yet.”

At the medical station, Van recorded An's breathing rate. He had no symptoms for the first ten minutes. After a truck passed by, his chest felt heavier even though the station was empty increase spray. Tung ran out to check. The car engine creates a clear jet of diesel, but the wind blows it into the measuring area and then separates. Huyen records traffic events in the context column.

A resident asked why it was necessary to record complicated experiments. Huyen replied because life does not separate variables for them. If they ditch the truck, the heat, the jogging, the stress and the food, they can create a beautiful result but it doesn't help anyone know what to do on the street.

Quynh walked between tables, listening more than talking. A worker gave her a form that said she had a headache but had no smell. Quynh asked him if he used a mask. He said yes, but removed it when the area “had been treated.” Quynh does not blame him. She added a sentence to the procedure sheet: “Odor reduction is not a signal to remove protection.”

At 4:12 p.m., a spray dose increased due to sensor fluctuations. Tung detected a decrease in the flow of the third nozzle. Huyen asked to remain active, not to interfere, to view the data. Tung objected because he did not want to turn people into test subjects.

“They agreed to participate,” Huyen said.

“Agreeing to participate does not mean consenting to exposure.”

Huyen stopped. Tung is right. She asked Quynh to temporarily stop the spraying, inform her of the reason, and then switch the test to the air sample she had taken before. Some people are frustrated that there isn't enough data yet. Huyen noted both the early stopping and the reasons related to the ethics of the trial.

Tu stood next to the table and gave Huyen a card describing: "There are experiments that fail because they know how to stop."

Tung read it and asked if he wrote it himself.

“No. A worker has lost the smell of writing.”

Tung looked at Huyen. “Is this card included in the report?”

“If you credit the source of the description and don't turn it into a technical conclusion.”

“Then give it.”

After stopping spraying, the yard quickly smelled of wet soil and car oil. People reacted differently. Some people covered their noses, some said they finally knew the direction of the wind, some felt their headaches subsided. The data do not provide a universal answer, but it does suggest that loss of smell is not always a unique benefit.

Huyen gathers groups. Each group must present a result that makes them unhappy. Station one cannot separate the sample smell from the market smell. Station two lost data for three minutes because of low battery. Station three had someone running before completing the measurement. Station four has someone next to you marking what you said.

“Do these things make testing meaningless?” asked a student.

"No. It said it had to be redesigned next time," Huyen said.

Quynh, standing next to him, added: "If we only announce clean results, we will create a fake clean program."

At the end of the session, An looked at the map. The red area does not form a continuous patch. It's dots along the truck tracks, a cluster around the school gate, a dot at the fish sauce stall, and a few green areas between them. Huyen asked if he wanted to erase the unclear dots.

“No. Write question mark.”

Van holds the symptom card. “This sample is not enough to be conclusive, but it is enough to ask for three more nights of monitoring.”

Tung checked station number 3 again. The flow rate was better after replacing the calibration box, but there was still a small fluctuation when the temperature decreased. He recommends not allowing the system to increase the dose automatically in experimental frames without an observer.

Quynh agreed, then signed immediately on the notice board.

Huyen put the blank paper in the file along with the map and imperfect forms. She wrote a preliminary conclusion:

“Testing the work plus the substance causing the symptoms cannot be identified. It determined that smell, machine and body did not change at the same time. Need to track by minute and by location.”

It was dark, people left. Nghi cleaned up the fish sauce jar, Tu collected the card, Van put away the symptom sheet, Tung locked the station. An stopped and looked at the cultural yard.

“Does the city understand?” he asked.

Huyen looked at the boards still covered with tape. “Don't know yet.”

An smiled. “That sentence sounds less scary when someone else does it.”

Huyen also laughed. During the night, the test data will be re-entered, checked, flagged and possibly partially rejected. It's okay. A city that learns to listen doesn't start with the right answer. It starts with not deleting things that make meetings more difficult.

After everyone left, Huyen and Diem stayed to check the handwritten votes. Two people received the same scent sample but used different words. Diem said that at first she wanted to follow Huyen's example, then remembered that the panel was not a competition with the same answers.

“If two people smell the same but smell differently, is the data weak?” Diem asked.

“It indicates a heterogeneous experience. We need to look at the context, not choose a speaker who suits us.”

Diem pointed to a slip that said there was no odor but her throat was dry. “I once abandoned this type because I thought the tester did not understand the task.”

Huyen said: "I was the same way before."

Two people put their votes in separate groups. Tung sent a report that injector number three was misaligned but no longer leaking. Van adds a column of time when symptoms decrease after leaving the area. Nghi called to ask if the control vial needed to be kept cold. Tu reminded them to write down anyone who did not participate because they could not smell.

Huyen realized the test did not end when people left the field. It continues in the way the team handles the excess, the fuzzy, the unstackable. She set up a short meeting the next morning just to ask: “What data did we throw away because we didn't know how to name it?”

Quynh immediately answered that it was the most time-consuming question. Huyen said maybe because it holds the most people. In the storage room, imperfect slips sit next to clear slips. No vote should be considered trash simply because it makes the final table difficult to see.